More fib plus pav updates

[dougollerenshaw]

Summary of updates:

• Add behavior cameras to src/aind_metadata_mapper/pavlovian_behavior/example_create_session.py
• Improves the docstring in src/aind_metadata_mapper/pavlovian_behavior/example_create_session.py with an example CLI command.
• Add a special case to src/aind_metadata_mapper/utils/merge_sessions.py to deal with conflicting reward units. If one session file has reward_consumed_total = null, ignore the associated units.
• Added many missing fields in the stimulus_epochs field in src/aind_metadata_mapper/pavlovian_behavior/example_create_session.py
• Vastly simplified scripts/example_create_fiber_and_pav_sessions.py to use the default params specified in src/aind_metadata_mapper/pavlovian_behavior/example_create_session.py and src/aind_metadata_mapper/fib/example_create_session.py
• Added ability to pass through the following parameters from the command line:
  • active_mouse_platform
  • anaesthesia
  • animal_weight_post
  • animal_weight_prior
  • mouse_platform_name
• Fixed incorrect timestamp format in session.json (timestamps should be in local timezone with offset, not in "Z" format)
• Added a new "timing_utils" module to handle shared utility functions for both fib and pavlovian_behavior mappers
• Updated the merge_sessions logic & added to the docstring to note that concept of merging sessions will go away w/ schema 2.0.

[dougollerenshaw]

Note: this is related to this issue.

[hagikent]

All changes confirmed. Major modality specific functions tested. Bash argparse example usage tested.
Once Saskia et al. give a green light to the (tentative) design of stimulus_epoch, goo to be merged! (or merge now and later patch-fix)

[dougollerenshaw]

Comment from Saskia: remove active_mouse_platform and anesthesia as passable params. They are not relevant here.

[dougollerenshaw]

@hagikent, I reviewed the session.json that is produced by this code with @saskiad yesterday afternoon. She had a few questions/suggestions. Here's a subset of those that I could use some help from you in answering:

• Are there any data acquisition boards you can specify that are used for behavior videos, behavior, or fiber? All of these are currently empty lists. They should correspond to daq names that are listed in the rig.json, which are:
  • "harp behavior board"
  • "harp sound card"
  • "harp clock synchronization board"
  • "harp sound amplifier"
  • "USB DAQ"
• The data_stream that I have named 'behavior' should actually be 'behavior-videos'. I'll make that change. But do you have any info you can share about the software that controlled the behavior videos? Ideally a github permalink to the actual version used.
• Do you have info on the software controlling the fiber acquisition? Again, ideally a github permalink to the actual version used
• And same for the behavior control. Need a link to the software used to control stim delivery
• Do the stimulus_names of the four stimuli in the stimulus_parameters (CS1, CS2, CS3, CS4) match the actual names used in code? If so, we should leave them as is. If not, Saskia was suggesting that the white noise stimulus should have a more descriptive name.

[hagikent]

1. not really — videos are digitized at the level of the sensors and directly USB connected to the PC. Fiber is the same. Behavior...? Not sure if any specific behavior, but mouse lick is digitized by lick-detector (analog to TTL) or "harp behavior board" (TTL to harp-message), depending on the definition of DAQ.

2. The same Bonsai workflow as the behavior control and FIP data acquisition.
   https://github.com/AllenNeuralDynamics/PavlovianCond_Bonsai

3. Same as above (all integrated)

4. Same as above (all integrated)

5. We need those abstraction since mapping can be changed (technically CS1 can be WN instead of CS4, although we haven't done such contingency flip yet)

[dougollerenshaw]

For software, I'm going to use the permalink for the version of the code after the most recent commit, which would be https://github.com/AllenNeuralDynamics/PavlovianCond_Bonsai/tree/dafd7dfe0f347f781e91466b3d16b83cf32f8b6d

[saskiad]

Went over output with Doug which looks good. Haven't examine code rigorously.

[jtyoung84]

I tend to discourage the use of writing to the user's file system during unit tests. For reading small files, we've been putting pre-made files into the test resources directory. For writes, we usually mock the write operation.

[dougollerenshaw]

Fixed in 5dbce34!

[jtyoung84]

• If possible, please update the unit tests to mock the write operations
• Also, in the future, please create a ticket with the proposed changes/issues before large modifications

[dougollerenshaw]

Thanks @jtyoung84!

• If possible, please update the unit tests to mock the write operations

Fixed.

• Also, in the future, please create a ticket with the proposed changes/issues before large modifications

Will do.

Requesting re-review.